Clinical utility of a liquid biopsy diagnostic approach in lung cancer using amplicon-based next-generation sequencing in parallel with allele-specific PCR.

Abstract

e21516 Background: The utility of a liquid biopsy test for diagnosis and treatment is determined by its accuracy, molecular target coverage and timeliness to a clinically informative result. We describe our clinical experience of providing comprehensive next-generation sequencing (NGS) plasma testing for advanced lung cancer cases, including parallel testing with allele-specific PCR (AS-PCR) for rapid EGFR mutation detection. Methods: Plasma cell-free DNA (cfDNA) from advanced lung cancer patients (n = 374) underwent real-world testing with an amplicon-based NGS assay, in a CAP and ISO15189 accredited central laboratory. The assay covers 51 genes, including 8 guideline recommended biomarkers- EGFR, BRAF, MET, ALK, RET, ROS1, ERBB2 and KRAS. 168 cases (44.9%) were treatment-naïve (baseline) and 206 cases (55.1%) had received one or more lines of treatment. Parallel testing with AS-PCR for 10 specific EGFR mutations was done for 151 cases (90 baseline, 61 non-baseline). Concordance of EGFR mutation detection by the two methods, and the frequency of detection of clinically actionable (driver) mutations by NGS were assessed. Turnaround time (TAT) was calculated from sample receipt. Results: An overall concordance of 97.4% was observed for EGFR mutations between AS-PCR and NGS. When restricted to baseline cases, concordance was 100%. Among baseline cases which were concordant with AS-PCR for EGFR negativity (n = 63), driver mutations were identified in EGFR (n = 7 rarer mutations), KRAS, ERBB2, MET, NRAS, BRAF and ALK (n = 16 total) by NGS, providing an additional 37% diagnostic yield (23/63 cases). Among all baseline cases tested by NGS, a driver gene mutation was found in 64.28% of cases, including EGFR (36.9%) and KRAS mutations (11.9 %). Among non-baseline cases which were EGFR-negative by NGS and by AS-PCR, added diagnostic yield provided by NGS was 33.3% (10 of 30 cases), and included 16.7% rarer EGFR mutations. Overall, detection rate of ALK, RET, ROS1 fusions was 2.4% (n = 5, 3 and 1, respectively). Median TAT for EGFR results by AS-PCR was 1 day (range 1-2 days), while median TAT for NGS results was 10 days (range 4-13 days). Conclusions: We report excellent real-world performance of blood-based liquid biopsy testing for detecting recommended biomarkers in lung cancers, including an approach combining two orthogonal platforms for quick decision making. Clinically meaningful diagnostic yields can be obtained using a timely comprehensive NGS assay, either individually or in parallel with rapid AS-PCR testing.